Less-supervised learning with knowledge distillation for sperm morphology analysis
Sperm Morphology Analysis (SMA) is pivotal in diagnosing male infertility. However, manual analysis is subjective and time-intensive. Artificial intelligence presents automated alternatives, but hurdles like limited data and image quality constraints hinder its efficacy. These challenges impede Deep...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Taylor & Francis Group
2024-12-01
|
| Series: | Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization |
| Subjects: | |
| Online Access: | https://www.tandfonline.com/doi/10.1080/21681163.2024.2347978 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1839635997343612928 |
|---|---|
| author | Ali Nabipour Mohammad Javad Shams Nejati Yasaman Boreshban Seyed Abolghasem Mirroshandel |
| author_facet | Ali Nabipour Mohammad Javad Shams Nejati Yasaman Boreshban Seyed Abolghasem Mirroshandel |
| author_sort | Ali Nabipour |
| collection | DOAJ |
| description | Sperm Morphology Analysis (SMA) is pivotal in diagnosing male infertility. However, manual analysis is subjective and time-intensive. Artificial intelligence presents automated alternatives, but hurdles like limited data and image quality constraints hinder its efficacy. These challenges impede Deep Learning (DL) models from grasping crucial sperm features. A solution enabling DL models to learn sample nuances, even with limited data, would be invaluable. This study proposes a Knowledge Distillation (KD) method to distinguish normal from abnormal sperm cells, leveraging the Modified Human Sperm Morphology Analysis dataset. Despite low-resolution, blurry images, our method yields relevant results. We exclusively utilize normal samples to train the model for anomaly detection, crucial in scenarios lacking abnormal data – a common issue in medical tasks. Our aim is to train an Anomaly Detection model using a dataset comprising unclear images and limited samples, without direct exposure to abnormal data. Our method achieves Receiver ROC/AUC scores of 70.4%, 87.6%, and 71.1% for head, vacuole, and acrosome, respectively, our method matches traditional DL model performance with less than 70% of the data. This less-supervised approach shows promise in advancing SMA despite data scarcity. Furthermore, KD enables model adaptability to edge devices in fertility clinics, requiring less processing power. |
| format | Article |
| id | doaj-art-25bbffd3f7e543fbb290a7a03ae2f44b |
| institution | Matheson Library |
| issn | 2168-1163 2168-1171 |
| language | English |
| publishDate | 2024-12-01 |
| publisher | Taylor & Francis Group |
| record_format | Article |
| series | Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization |
| spelling | doaj-art-25bbffd3f7e543fbb290a7a03ae2f44b2025-07-08T10:28:41ZengTaylor & Francis GroupComputer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization2168-11632168-11712024-12-0112110.1080/21681163.2024.2347978Less-supervised learning with knowledge distillation for sperm morphology analysisAli Nabipour0Mohammad Javad Shams Nejati1Yasaman Boreshban2Seyed Abolghasem Mirroshandel3Department of Computer Engineering, Faculty of Engineering, University of Guilan, Rasht, IranDepartment of Computer Engineering, Faculty of Engineering, University of Guilan, Rasht, IranDepartment of Computer Engineering, Faculty of Engineering, University of Guilan, Rasht, IranDepartment of Computer Engineering, Faculty of Engineering, University of Guilan, Rasht, IranSperm Morphology Analysis (SMA) is pivotal in diagnosing male infertility. However, manual analysis is subjective and time-intensive. Artificial intelligence presents automated alternatives, but hurdles like limited data and image quality constraints hinder its efficacy. These challenges impede Deep Learning (DL) models from grasping crucial sperm features. A solution enabling DL models to learn sample nuances, even with limited data, would be invaluable. This study proposes a Knowledge Distillation (KD) method to distinguish normal from abnormal sperm cells, leveraging the Modified Human Sperm Morphology Analysis dataset. Despite low-resolution, blurry images, our method yields relevant results. We exclusively utilize normal samples to train the model for anomaly detection, crucial in scenarios lacking abnormal data – a common issue in medical tasks. Our aim is to train an Anomaly Detection model using a dataset comprising unclear images and limited samples, without direct exposure to abnormal data. Our method achieves Receiver ROC/AUC scores of 70.4%, 87.6%, and 71.1% for head, vacuole, and acrosome, respectively, our method matches traditional DL model performance with less than 70% of the data. This less-supervised approach shows promise in advancing SMA despite data scarcity. Furthermore, KD enables model adaptability to edge devices in fertility clinics, requiring less processing power.https://www.tandfonline.com/doi/10.1080/21681163.2024.2347978Human sperm morphometrysperm defectsinfertilitydeep learningknowledge distillation |
| spellingShingle | Ali Nabipour Mohammad Javad Shams Nejati Yasaman Boreshban Seyed Abolghasem Mirroshandel Less-supervised learning with knowledge distillation for sperm morphology analysis Computer Methods in Biomechanics and Biomedical Engineering: Imaging & Visualization Human sperm morphometry sperm defects infertility deep learning knowledge distillation |
| title | Less-supervised learning with knowledge distillation for sperm morphology analysis |
| title_full | Less-supervised learning with knowledge distillation for sperm morphology analysis |
| title_fullStr | Less-supervised learning with knowledge distillation for sperm morphology analysis |
| title_full_unstemmed | Less-supervised learning with knowledge distillation for sperm morphology analysis |
| title_short | Less-supervised learning with knowledge distillation for sperm morphology analysis |
| title_sort | less supervised learning with knowledge distillation for sperm morphology analysis |
| topic | Human sperm morphometry sperm defects infertility deep learning knowledge distillation |
| url | https://www.tandfonline.com/doi/10.1080/21681163.2024.2347978 |
| work_keys_str_mv | AT alinabipour lesssupervisedlearningwithknowledgedistillationforspermmorphologyanalysis AT mohammadjavadshamsnejati lesssupervisedlearningwithknowledgedistillationforspermmorphologyanalysis AT yasamanboreshban lesssupervisedlearningwithknowledgedistillationforspermmorphologyanalysis AT seyedabolghasemmirroshandel lesssupervisedlearningwithknowledgedistillationforspermmorphologyanalysis |